Nielsen is making changes to its data measurement heading into the new television season, headlined by passive measurement of co-viewing becoming part of its national currency.
Nielsen CEO Karthik Rao said in a statement that the company has spent “months working hand in hand” with its clients and industry figures “to make Big Data + Panel even more accurate.”
The change will impact the 40 million homes under the Big Data + Panel measurement, plus an additional 15,000 homes exclusively with set meters with plans to add people-meter data in the future. While co-viewing was measured initially through people meters, this new approach does not require a formal login process and uses the company’s wearable devices to account for the measurement.
“The difference is this approach is passive,” Brian Fuhrer, SVP/product strategy and thought leadership at Nielsen, said in an interview with Sports Media Watch Tuesday. “As you know, people tend to be inherently lazy, so when they have the paradox of choice of, ‘Do I do anything or do nothing?,’ they tend to do nothing. So we have to overcome that paradox by urging them to comply because of the importance of being in the panel. That, by the way, is the No. 1 reason why homes are only in the panel for two years — that fatigue of pushing buttons [is] the key thing that we focus on.”
Nielsen ran a co-viewing pilot program in February and found that “marquee live events,” such as the Super Bowl, NBA All-Star Game and the opening and closing ceremonies of the Olympics, experienced a 4.19% average viewership lift. While those events have particularly high degrees of co-viewing, the company expects some level of normalization as the measurement system is used on more ordinary events.
NFL chief data and analytics officer Paul Ballew said in an interview with Joe Flint of The Wall Street Journal before last season that Nielsen was undercounting co-viewing for its games, adding during a later conference call with reporters that the league believed “very strongly there’s more work to be done.” Ballew, then: “They know our ongoing passion that co-viewing is inaccurately measured, especially for big events. If you think about the Super Bowl, for instance, where they’re showing a co-viewing factor of 2.4, it just makes no sense and doesn’t tie to other work we’ve done on that front.” The difference, he said, could represent 15-20 million additional viewers for the game.
When asked about the impetus for the co-viewing changes, Fuhrer said that Nielsen looked at the metric after reviewing its own data sets and the external marketplace.
“If you look across the industry, there’s a lot of concern or suggestion that there’s an opportunity to get more information about the people, and clearly, we’ve been looking at trends within our own panel,” Fuhrer said. “We look at a long-term trend of, ‘What is the household rating?,’ and, ‘What is the persons rating?,’ and then, ‘What does each demographic group look like?,’ and, ‘How do those trend-lines compare?,’ and as we saw these kind of coming together, it was clear that there was an opportunity to improve the instrument.”
Nielsen is also fully integrating latency adjusted DASH universe estimates into its measurements, primarily impacting streaming viewership. The company already rolled out the new universe estimate system in February, but it pulled it back from The Gauge amid drastic trend changes. Streaming viewership is expected to have a minor decline with this change, while cable would stay flat or have a slight increase.
“Some of the latency amplification that there was with DASH is going to be muted somewhat because it’s going to be more recent data with cable,” Fuhrer said. “Obviously, cable’s trending downward, but the co-viewing component should help in that regard, so there’s going to be a lot of offsets, but in most cases, we’re seeing flat to slight increases.”
Nielsen has included a note within its Gauge chart, noting that the measurements “do not reflect Nielsen’s currency TV ratings.” The June edition of the chart showed streaming at a 48.5% share of P2+ total day television viewing, while cable and broadcast combine for a 39.3% share. DASH UE, which is derived through a partnership between the Advertising Research Foundation and NORC at the University of Chicago, is accredited by the Media Rating Council. Whereas Nielsen used to update its universe once per year, it is now doing so more frequently with changes to the media ecosystem.
Nielsen is introducing a series of additional changes related to demographics, automated content recognition and household information. The company is also refining how it weights its measurements from Big Data and panel-only data. When the company first implemented the Big Data + Panel system, it altered weighting with a focus on linear television. Fuhrer said that the company made adjustments to weighting based on demographic groups within Big Data. The change will now better encompass a sample that includes streaming television as well, providing a more complete perspective on viewership.
“I don’t think there’s going to be any big changes to the totals or anything like that,” Fuhrer said, “but I think overall, it’s going to be more consistent and stable across the demographic groups.”
Nielsen has made several changes to its viewership measurements over the years, the most recent of which have been partially attributed to a rise in sports viewership. Most media companies and leagues are reporting Big Data + Panel, which is considered currency. It is official Nielsen policy to compare the Big Data + Panel metrics to panel-only measurements from the previous year.
“There’s always things that we need to adjust and change,” Fuhrer said. “We’re going to have more opportunity to integrate different data sets and improve the demographic models, and different sources for universe estimates are going to be made available, so it’s just a constant work in progress.”
